Designing a Monitoring Study
Learning Objectives
Students will be able to:
- Formulate a testable research question about indoor air quality
- Identify independent, dependent, and controlled variables
- Design a data collection protocol with appropriate sampling
- Anticipate potential sources of error and bias
From Question to Study
Good science starts with good questions.
Now that you understand how sensors work and what they measure, it's time to design your own investigation. A well-designed study helps you collect data that actually answers your question.
Step 1: Choose Your Research Question
A good research question is:
- Specific: Focused enough to answer with available tools
- Measurable: Can be answered with sensor data
- Achievable: Possible within time and resource constraints
- Relevant: Meaningful to you and your community
Example Research Questions
Good Questions
- Does CO2 increase during class vs. passing periods?
- How does opening windows affect CO2 levels?
- Is PM2.5 higher near the cafeteria during lunch?
- How do CO2 levels compare in different classrooms?
Needs Improvement
- "Is the air quality good?" (not specific)
- "Why is pollution bad?" (not measurable)
- "What's the air like in China?" (not achievable)
- "What was air like 100 years ago?" (can't measure)
Step 2: Identify Your Variables
Independent Variable
What you change or compare
Examples: window open vs. closed, time of day, room location, number of people
Dependent Variable
What you measure
Examples: CO2 concentration (ppm), PM2.5 level (μg/m³), temperature
Controlled Variables
What you keep the same
Examples: same sensor, same location in room, same time of day, same weather conditions
Example: "Does opening windows lower CO2?"
| Independent: | Window status (open vs. closed) |
| Dependent: | CO2 level in ppm |
| Controlled: | Same room, same number of people, same time of day, same sensor position |
Step 3: Plan Your Sampling
When to Measure
- Frequency: How often? (every minute, every 5 minutes, hourly)
- Duration: How long total? (one class, one day, one week)
- Timing: At what times? (morning, afternoon, specific events)
- Replication: How many days/sessions?
Where to Measure
- Location in room: Center? Near window? Near door?
- Height: Breathing zone (~4 feet) is often best
- Distance from sources: Away from vents, not near doors
- Multiple locations? Compare different spots
Step 4: Create a Data Collection Protocol
A protocol is a step-by-step procedure that ensures consistency.
Sample Protocol
Study: Effect of Class Size on CO2
- Place sensor in center of room, 4 feet high, at least 3 feet from any person
- Allow sensor to stabilize for 5 minutes before recording
- Record reading every 5 minutes for the entire class period
- Note the time, reading, number of students, and window/door status
- Record any unusual events (door opened, someone coughed near sensor, etc.)
- At end of class, count total number of people present
- Repeat for at least 3 different class periods
Step 5: Consider Sources of Error
Common Errors
- Sensor not calibrated correctly
- Measuring too close to a source
- Not waiting for sensor to stabilize
- Inconsistent timing between readings
- Changes in conditions you didn't account for
How to Minimize Errors
- Follow your protocol exactly every time
- Record any deviations or unusual events
- Take multiple measurements
- Use the same sensor throughout
- Control as many variables as possible
Activity: Design Your Study
Study Design Worksheet
Work in groups to complete your study design:
| 1. Research Question: | |
| 2. Independent Variable: | |
| 3. Dependent Variable: | |
| 4. Controlled Variables: | |
| 5. Measurement Frequency: | |
| 6. Total Duration: | |
| 7. Sensor Location: | |
| 8. Potential Errors: |
Peer Review
Exchange designs with another group. Check: Is the question specific? Are variables clearly identified? Is the protocol detailed enough to replicate?
Key Takeaway
A well-designed study starts with a specific, measurable question and clearly identifies what you're changing, what you're measuring, and what you're keeping the same. A detailed protocol ensures that your data collection is consistent and that someone else could replicate your study. Good planning now means better data later!